In this context, we now report the generation and reactivity
with oxygen of the radicals derived from isobenzofuranone
compounds in Scheme 1. We selected these molecules to
Table 1. Absorption of the Radicals, Rate Constants for the
Reactions of Cumyloxyl/tert-Butoxyl Radicals with Hydrogen
Donors, and Reactivity toward Oxygen
alkoxyl radical
rate constant
radical
wavelength
max, nm
reactivity
with
oxygena
Scheme 1. Isobenzofuranones Studied and Related Compounds
substrate
(106 M-1 s-1
)
HP-1364
12.4b
51c
340
400
360, 525
360
340
340, 380
340, 580
340
no
no
2-coumaranone4
1
2
3
4
5
6
7
2.07c
yes
(no)d
yes
yes
no
5.35e
2.83e
3.77e
3.01e
2.70e
yes
yes
340
a Based of laser flash photolysis work in a 100-µs time scale. b Obtained
by laser flash photolysis of dicumyl peroxide. c Obtained by laser flash
photolysis of di-tert-butyl peroxide. d This measurements are difficult due
to potential interference of Br2•- at 360 nm. There are also poorly resolved
absorbances in the 500-600 nm range; these are normal for a radical with
this structure. e Obtained from product studies with cumyl hyponitrite as
the source of cumyloxyl radicals.
bands were detectable between 300 and 700 nm, indicating
a low contribution of the carbonyl group to the absorption
of the radical. Further, the signal for the radical from 3 was
quenched by oxygen with a rate constant of 7.2 × 108 M-1
s-1. This is about a million times faster than the isomeric
radical from 2-coumaranone.
Benzyl radicals are readily quenched by paramagnetic
substrates such as nitroxide radicals (the rate constants for
these reactions are frequently ∼108 M-1 s-1 in acetonitrile).9
Thus, the rate constants of the radicals generated from 3 and
the HP-136 toward 2,2,5,5-tetramethylpiperidin-1-oxyl
(TEMPO) were obtained from the transient decay at different
TEMPO concentrations in mixtures of benzene/di-tert-butyl
peroxide (1/1 v/v). The values obtained are 1.77 × 107 and
1.35 × 105 M-1 s-1, respectively. Clearly, the benzofuranone
structure leads to a more stable, and therefore less reactive
radical.
From the above results, it seems that the contribution of
the resonance form centered on oxygen is decisive to explain
the lack of reactivity of the benzofuranone structure toward
oxygen. In fact, spin density calculation data10 for the radical
from phthalide 1 show higher spin density located on the
benzylic carbon atom (52.4%) than for the radical from
2-coumaranone (45.3%).
compare their reactivity toward oxygen with those previously
reported for benzofuranones4 and to establish the importance
of electron-withdrawing effects due to the presence of nitro
or trifluoromethyl groups.
The radicals were generated from the corresponding
precursors by laser excitation of di-tert-butyl peroxide in
benzene 50% (v/v), using 355-nm laser pulses (radical
absorption maxima are given in Table 1). The tert-butoxyl
radicals abstract the benzylic hydrogen to produce carbon-
centered radicals and tert-butyl alcohol. The growth of the
radical signal reflects the H-abstraction and other forms of
decay of tert-butoxyl such as reaction with the solvent and
â-cleavage.7
Under a nitrogen atmosphere the radical generated from
3H-isobenzofuran-1-one (1) absorbs at 360 and 525 nm. The
visible band is more intense than in the case of the benzyl
radical; this enhancement is probably due to the presence of
heteroatoms. A similar behavior has been previously de-
scribed for the 2-coumaranone radical; however, in this case
the maximum was detected at 400 nm and the intensity of
the band was far greater than that in the case of the radical
from 1. When the same experiments are carried out in
oxygen-saturated samples the transient is totally quenched.
The photolysis of di-tert-butyl peroxide under nitrogen in
the presence of 3-phenyl-3H-isobenzofuran-1-one (3) (0.05
M) yields a transient with maximum at 340 nm; this band is
quite similar to that for the diphenylmethyl radical.8 No other
If we assume that the spin density at the benzylic position
is key to the coupling reaction of benzyl radical and
(8) Scaiano, J. C.; Tanner, M.; Weir, D. J. Am. Chem. Soc. 1985, 107,
4396.
(9) Beckwith, A. L. J.; Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc.
1992, 114, 4983.
(10) Spin density calculations were calculated by Density Functional
Theory. Electronic structures of the radical molecules were calculated with
use of the Spartan software package (V.5.0.3, Wavefunction, Inc. Irvine).
The molecular structures were optimized employing a split-valence basis
with a single set of polarization functions added to each atom (DN**). A
gradient-corrected functional combining Becke-Perdew as the exchange-
correlation functional (BP86) was chosen. For more details see the
Supporting Information.
(7) Paul, H.; Small, R. D.; Scaiano, J. C. J. Am. Chem. Soc. 1978, 100,
4520.
1516
Org. Lett., Vol. 5, No. 9, 2003